def train(dir):
"""
Command to train ML model(s) locally
"""
logger.info(ASCII_LOGO)
logger.info("Started local training...\n")
sagify_module_path = os.path.join(dir, 'sagify')
local_train_script_path = os.path.join(sagify_module_path, 'local_test',
'train_local.sh')
test_path = os.path.join(sagify_module_path, 'local_test', 'test_dir')
if not os.path.isdir(test_path):
logger.info("This is not a sagify directory: {}".format(dir))
sys.exit(-1)
try:
subprocess.check_output([
"{}".format(local_train_script_path),
"{}".format(os.path.abspath(test_path))
])
logger.info("Local training completed successfully!")
except Exception as e:
logger.info("{}".format(e))
return
def _run_test_script(self,
filename='mytestscript.py',
interpreter=sys.executable):
# Absolute file path:
fn = self.tempdir + filename
try:
output = check_output([interpreter, fn],
env=self.env,
stderr=STDOUT)
except CalledProcessError as e:
with open(fn) as f:
msg = ('Error running the command %s\n'
'%s\n'
'Contents of file %s:\n'
'\n'
'%s') % (
' '.join([interpreter, fn]),
'env=%s' % self.env,
fn,
'----\n%s\n----' % f.read(),
)
if not hasattr(e, 'output'):
# The attribute CalledProcessError.output doesn't exist on Py2.6
e.output = None
raise VerboseCalledProcessError(msg,
e.returncode,
e.cmd,
output=e.output)
return output
def build(dir, requirements_dir, docker_tag):
"""
Builds a Docker image that contains code under the given source root directory.
Assumes that Docker is installed and running locally.
:param dir: [str], source root directory
:param requirements_dir: [str], path to requirements.txt
"""
sagify_module_path = os.path.relpath(os.path.join(dir, 'sagify/'))
build_script_path = os.path.join(sagify_module_path, 'build.sh')
dockerfile_path = os.path.join(sagify_module_path, 'Dockerfile')
train_file_path = os.path.join(sagify_module_path, 'training', 'train')
serve_file_path = os.path.join(sagify_module_path, 'prediction', 'serve')
executor_file_path = os.path.join(sagify_module_path, 'executor.sh')
if not os.path.isfile(build_script_path) or not os.path.isfile(train_file_path) or not \
os.path.isfile(serve_file_path):
raise ValueError("This is not a sagify directory: {}".format(dir))
os.chmod(train_file_path, 0o777)
os.chmod(serve_file_path, 0o777)
os.chmod(executor_file_path, 0o777)
target_dir_name = os.path.basename(os.path.normpath(dir))
output = subprocess.check_output([
"{}".format(build_script_path), "{}".format(os.path.relpath(dir)),
"{}".format(os.path.relpath(target_dir_name)),
"{}".format(dockerfile_path),
"{}".format(os.path.relpath(requirements_dir)), docker_tag
])
logger.debug(output)
def push(dir, docker_tag, aws_region, iam_role_arn, aws_profile, external_id,
image_name):
"""
Push Docker image to AWS ECS
:param dir: [str], source root directory
:param docker_tag: [str], the Docker tag for the image
:param aws_region: [str], the AWS region to push the image to
:param iam_role_arn: [str], the AWS role used to push the image to ECR
:param aws_profile: [str], the AWS profile used to push the image to ECR
:param external_id: [str], Optional external id used when using an IAM role
:param image_name: [str], The name of the Docker image
"""
sagify_module_path = os.path.relpath(os.path.join(dir, 'sagify/'))
push_script_path = os.path.join(sagify_module_path, 'push.sh')
if not os.path.isfile(push_script_path):
raise ValueError("This is not a sagify directory: {}".format(dir))
output = subprocess.check_output([
"{}".format(push_script_path), docker_tag, aws_region, iam_role_arn,
aws_profile, external_id, image_name
])
logger.debug(output)
def _run_test_script(self, filename='mytestscript.py',
interpreter=sys.executable):
# Absolute file path:
fn = self.tempdir + filename
try:
output = check_output([interpreter, fn],
env=self.env, stderr=STDOUT)
except CalledProcessError as e:
with open(fn) as f:
msg = (
'Error running the command %s\n'
'%s\n'
'Contents of file %s:\n'
'\n'
'%s') % (
' '.join([interpreter, fn]),
'env=%s' % self.env,
fn,
'----\n%s\n----' % f.read(),
)
if not hasattr(e, 'output'):
# The attribute CalledProcessError.output doesn't exist on Py2.6
e.output = None
raise VerboseCalledProcessError(msg, e.returncode, e.cmd, output=e.output)
return output
def download_apk(storage, handle, tmp_dir, apk_name, apk_tmp):
"""
Download the APK from Google Play for the given handle.
:param storage: minio storage helper
:param handle: application handle to download
:param tmp_dir: download destination directory
:param apk_name: name of the APK in Minio storage
:param apk_tmp: apk temporary name
:return: True if succeed, False otherwise
"""
DEVICE_CODE_NAMES = [
'bacon', 'hammerhead', 'manta', 'cloudbook', 'bullhead'
]
RETRY_PER_DEVICE = 3
refreshed_token = False
for device in DEVICE_CODE_NAMES:
cmd = 'gplaycli -v -a -y -pd {} -dc {} -f {}/'.format(
handle, device, tmp_dir)
for i in range(RETRY_PER_DEVICE):
try:
output = subprocess.check_output(
shlex.split(cmd),
stderr=subprocess.STDOUT,
timeout=240 # Timeout of 4 minutes
)
output_str = output.decode('utf-8')
logging.info(output_str)
if '[ERROR]' in output_str:
filtered = output_str.replace(
'[ERROR] cache file does not exists or is corrupted',
'')
if '[ERROR]' in filtered:
raise RuntimeError('Error while downloading apk file')
apk = Path(apk_tmp)
if apk.is_file():
try:
storage.put_file(apk_tmp, apk_name)
return True
except ResponseError as err:
logging.info(err)
return False
except TimeoutExpired:
logging.warning("Timeout of gplaycli download command")
return False
except Exception as e:
logging.info(e)
logging.info("Could not download with device {}".format(device))
if not refreshed_token:
remove_token()
refreshed_token = True
time.sleep(2)
return False
def build(dir, requirements_dir):
"""
Command to build SageMaker app
"""
logger.info(ASCII_LOGO)
logger.info(
"Started building SageMaker Docker image. It will take some minutes...\n"
)
sagify_module_path = os.path.relpath(os.path.join(dir, 'sagify/'))
build_script_path = os.path.join(sagify_module_path, 'build.sh')
dockerfile_path = os.path.join(sagify_module_path, 'Dockerfile')
train_file_path = os.path.join(sagify_module_path, 'training', 'train')
serve_file_path = os.path.join(sagify_module_path, 'prediction', 'serve')
executor_file_path = os.path.join(sagify_module_path, 'executor.sh')
if not os.path.isfile(build_script_path) or not os.path.isfile(train_file_path) or not \
os.path.isfile(serve_file_path):
logger.info("This is not a sagify directory: {}".format(dir))
sys.exit(-1)
os.chmod(train_file_path, 0o777)
os.chmod(serve_file_path, 0o777)
os.chmod(executor_file_path, 0o777)
target_dir_name = os.path.basename(os.path.normpath(dir))
try:
subprocess.check_output([
"{}".format(build_script_path), "{}".format(os.path.relpath(dir)),
"{}".format(os.path.relpath(target_dir_name)),
"{}".format(dockerfile_path),
"{}".format(os.path.relpath(requirements_dir))
])
logger.info("Docker image built successfully!")
except Exception as e:
logger.info("{}".format(e))
return
def _futurize_test_script(self,
filename='mytestscript.py',
stages=(1, 2),
all_imports=False,
from3=False,
conservative=False):
params = []
stages = list(stages)
if all_imports:
params.append('--all-imports')
if from3:
script = 'pasteurize.py'
else:
script = 'futurize.py'
if stages == [1]:
params.append('--stage1')
elif stages == [2]:
params.append('--stage2')
else:
assert stages == [1, 2]
if conservative:
params.append('--conservative')
# No extra params needed
# Absolute file path:
fn = self.tempdir + filename
call_args = [sys.executable, script] + params + ['-w', fn]
try:
output = check_output(call_args, stderr=STDOUT, env=self.env)
except CalledProcessError as e:
with open(fn) as f:
msg = ('Error running the command %s\n'
'%s\n'
'Contents of file %s:\n'
'\n'
'%s') % (
' '.join(call_args),
'env=%s' % self.env,
fn,
'----\n%s\n----' % f.read(),
)
ErrorClass = (FuturizeError
if 'futurize' in script else PasteurizeError)
if not hasattr(e, 'output'):
# The attribute CalledProcessError.output doesn't exist on Py2.6
e.output = None
raise ErrorClass(msg, e.returncode, e.cmd, output=e.output)
return output
def push(dir):
"""
Command to push Docker image to AWS ECS
"""
logger.info(ASCII_LOGO)
logger.info(
"Started pushing Docker image to AWS ECS. It will take some time. Please, be patient...\n"
)
sagify_module_path = os.path.relpath(os.path.join(dir, 'sagify/'))
push_script_path = os.path.join(sagify_module_path, 'push.sh')
if not os.path.isfile(push_script_path):
logger.info("This is not a sagify directory: {}".format(dir))
sys.exit(-1)
try:
subprocess.check_output(["{}".format(push_script_path)])
logger.info("Docker image pushed to ECS successfully!")
except Exception as e:
logger.info("{}".format(e))
return
def push(dir, docker_tag):
"""
Push Docker image to AWS ECS
:param dir: [str], source root directory
"""
sagify_module_path = os.path.relpath(os.path.join(dir, 'sagify/'))
push_script_path = os.path.join(sagify_module_path, 'push.sh')
if not os.path.isfile(push_script_path):
raise ValueError("This is not a sagify directory: {}".format(dir))
output = subprocess.check_output(["{}".format(push_script_path), docker_tag])
logger.debug(output)
def deploy(dir):
"""
Command to deploy ML model(s) locally
"""
logger.info(ASCII_LOGO)
logger.info("Started local deployment at localhost:8080 ...\n")
sagify_module_path = os.path.join(dir, 'sagify')
local_deploy_script_path = os.path.join(sagify_module_path, 'local_test',
'deploy_local.sh')
test_path = os.path.join(sagify_module_path, 'local_test', 'test_dir')
if not os.path.isdir(test_path):
logger.info("This is not a sagify directory: {}".format(dir))
sys.exit(-1)
try:
subprocess.check_output([
"{}".format(local_deploy_script_path),
"{}".format(os.path.abspath(test_path))
])
except Exception as e:
logger.info("{}".format(e))
return
def load_kubeconf(config_file):
with open(config_file, 'r') as f:
config = yaml.load(f)
user = [user for user in config['users']
if user['name'] == 'aws'][0]['user']
command = [user['exec']['command']]
command.extend(user['exec']['args'])
output = subprocess.check_output(command)
c = json.loads(output.decode('utf-8'))
user['token'] = c['status']['token']
del user['exec']
loader = KubeConfigLoader(config)
config = type.__call__(Configuration)
loader.load_and_set(config)
Configuration.set_default(config)
def test_kafka_build_and_inject(tmpdir):
dest = test_kafka_build(tmpdir)
kafka_conf = str(check_output([
'env',
'ENV_INJECT=*.properties:/dev/null,server.properties:-',
'KAFKA_SERVER_ZOOKEEPER_CONNECT=zookeeper:2181/foo/bar',
'KAFKA_SERVER_ASDF=asdf',
'env2config',
'inject',
str(dest),
]))
assert 'replacing default' in kafka_conf
assert 'zookeeper.connect=zookeeper:2181/foo/bar' in kafka_conf
assert 'not matching any default' in kafka_conf
assert 'asdf=asdf' in kafka_conf
def _futurize_test_script(
self,
filename="mytestscript.py",
stages=(1, 2),
all_imports=False,
from3=False,
conservative=False,
):
params = []
stages = list(stages)
if all_imports:
params.append("--all-imports")
if from3:
script = "pasteurize.py"
else:
script = "futurize.py"
if stages == [1]:
params.append("--stage1")
elif stages == [2]:
params.append("--stage2")
else:
assert stages == [1, 2]
if conservative:
params.append("--conservative")
# No extra params needed
# Absolute file path:
fn = self.tempdir + filename
call_args = [sys.executable, script] + params + ["-w", fn]
try:
output = check_output(call_args, stderr=STDOUT, env=self.env)
except CalledProcessError as e:
with open(fn) as f:
msg = ("Error running the command %s\n"
"%s\n"
"Contents of file %s:\n"
"\n"
"%s") % (
" ".join(call_args),
"env=%s" % self.env,
fn,
"----\n%s\n----" % f.read(),
)
ErrorClass = FuturizeError if "futurize" in script else PasteurizeError
raise ErrorClass(msg, e.returncode, e.cmd, output=e.output)
return output
def _run_test_script(self, filename='mytestscript.py',
interpreter=sys.executable):
# Absolute file path:
fn = self.tempdir + filename
try:
output = check_output([interpreter, fn],
env=self.env, stderr=STDOUT)
except CalledProcessError as e:
msg = ('Error running the command %s\n%s\nContents of file %s:\n\n%s' %
(' '.join([interpreter, fn]),
'env=%s' % self.env,
fn,
'----\n%s\n----' % open(fn).read(),
)
)
raise VerboseCalledProcessError(msg, e.returncode, e.cmd, output=e.output)
return output
def test_redis_build_and_inject(tmpdir):
dest = test_redis_build(tmpdir)
redis_conf = str(check_output([
'env',
'ENV_INJECT=redis.conf:-',
'REDIS_APPENDONLY=yes',
'REDIS_ASDF=asdf',
'env2config',
'inject',
str(dest),
]))
assert 'replacing default' in redis_conf
assert 'appendonly yes' in redis_conf
assert 'not matching any default' in redis_conf
assert 'asdf asdf' in redis_conf
def test_redis_build_and_inject(tmpdir):
dest = test_redis_build(tmpdir)
redis_conf = str(
check_output([
'env',
'ENV_INJECT=redis.conf:-',
'REDIS_APPENDONLY=yes',
'REDIS_ASDF=asdf',
'env2config',
'inject',
str(dest),
]))
assert 'replacing default' in redis_conf
assert 'appendonly yes' in redis_conf
assert 'not matching any default' in redis_conf
assert 'asdf asdf' in redis_conf
def check_command_output(command, log, env=None, raise_on_error=True):
"""
Execute shell command and retrieve command output.
:param command: command to execute
:param env: a dictionary containing environment variables
:param log: logger
:param raise_on_error: True to raise subprocess.CalledProcessError on errors
:return: the command output
:raise: subprocess.CalledProcessError if the command fails
"""
return _run_command(
lambda _command, _env: check_output(_command, env=_env, stderr=subprocess.STDOUT, universal_newlines=True),
command,
log,
env,
raise_on_error,
)
def test_hadoop_build_and_inject(tmpdir):
dest = test_hadoop_build(tmpdir)
hadoop_confs = str(check_output([
'env',
'ENV_INJECT=*.xml:-',
'HADOOP_HDFS_DFS_NAMENODE_RPC-ADDRESS=wacky-namenode',
'HADOOP_YARN_YARN_RESOURCEMANAGER_ADDRESS=wacky-resman',
'HADOOP_MAPRED_MAPREDUCE_JOBTRACKER_JOBHISTORY_LOCATION=wacky-mrhistory',
'env2config',
'inject',
str(dest),
]))
assert 'wacky-namenode' in hadoop_confs
assert 'wacky-resman' in hadoop_confs
assert 'wacky-mrhistory' in hadoop_confs
assert len(list(re.finditer('Injected by env2config', hadoop_confs))) == 3
def check_command_output(command, env=None, raise_on_error=True):
"""
Execute shell command and retrieve command output.
:param command: command to execute
:param env: a dictionary containing environment variables
:param raise_on_error: True to raise subprocess.CalledProcessError on errors
:return: the command output
:raise: subprocess.CalledProcessError if the command fails
"""
return _run_command(
lambda _command, _env: check_output(_command,
env=_env,
stderr=subprocess.STDOUT,
universal_newlines=True),
command,
env,
raise_on_error,
)
def _futurize_test_script(self, filename='mytestscript.py', stages=(1, 2),
all_imports=False, from3=False,
conservative=False):
params = []
stages = list(stages)
if all_imports:
params.append('--all-imports')
if from3:
script = 'pasteurize.py'
else:
script = 'futurize.py'
if stages == [1]:
params.append('--stage1')
elif stages == [2]:
params.append('--stage2')
else:
assert stages == [1, 2]
if conservative:
params.append('--conservative')
# No extra params needed
# Absolute file path:
fn = self.tempdir + filename
call_args = [sys.executable, script] + params + ['-w', fn]
try:
output = check_output(call_args, stderr=STDOUT, env=self.env)
except CalledProcessError as e:
with open(fn) as f:
msg = (
'Error running the command %s\n'
'%s\n'
'Contents of file %s:\n'
'\n'
'%s') % (
' '.join(call_args),
'env=%s' % self.env,
fn,
'----\n%s\n----' % f.read(),
)
ErrorClass = (FuturizeError if 'futurize' in script else PasteurizeError)
raise ErrorClass(msg, e.returncode, e.cmd, output=e.output)
return output
def deploy(dir, docker_tag):
"""
Deploys ML models(s) locally
:param dir: [str], source root directory
:param docker_tag: [str], the Docker tag for the image
"""
sagify_module_path = os.path.join(dir, 'sagify')
local_deploy_script_path = os.path.join(sagify_module_path, 'local_test',
'deploy_local.sh')
test_path = os.path.join(sagify_module_path, 'local_test', 'test_dir')
if not os.path.isdir(test_path):
raise ValueError("This is not a sagify directory: {}".format(dir))
output = subprocess.check_output([
"{}".format(local_deploy_script_path),
"{}".format(os.path.abspath(test_path)), docker_tag
])
logger.debug(output)
def train(dir, docker_tag, image_name):
"""
Trains ML model(s) locally
:param dir: [str], source root directory
:param docker_tag: [str], the Docker tag for the image
:param image_name: [str], The name of the Docker image
"""
sagify_module_path = os.path.join(dir, 'sagify')
local_train_script_path = os.path.join(sagify_module_path, 'local_test',
'train_local.sh')
test_path = os.path.join(sagify_module_path, 'local_test', 'test_dir')
if not os.path.isdir(test_path):
raise ValueError("This is not a sagify directory: {}".format(dir))
output = subprocess.check_output([
"{}".format(local_train_script_path),
"{}".format(os.path.abspath(test_path)), docker_tag, image_name
])
logger.debug(output)
def test_hadoop_build_and_inject(tmpdir):
dest = test_hadoop_build(tmpdir)
hadoop_confs = str(
check_output(
[
"env",
"ENV_INJECT=*.xml:-",
"HADOOP_HDFS_DFS_NAMENODE_RPC-ADDRESS=wacky-namenode",
"HADOOP_YARN_YARN_RESOURCEMANAGER_ADDRESS=wacky-resman",
"HADOOP_MAPRED_MAPREDUCE_JOBTRACKER_JOBHISTORY_LOCATION=wacky-mrhistory",
"env2config",
"inject",
str(dest),
]
)
)
assert "wacky-namenode" in hadoop_confs
assert "wacky-resman" in hadoop_confs
assert "wacky-mrhistory" in hadoop_confs
assert len(list(re.finditer("Injected by env2config", hadoop_confs))) == 3
def _run_test_script(self,
filename='mytestscript.py',
interpreter=sys.executable):
# Absolute file path:
fn = self.tempdir + filename
try:
output = check_output([interpreter, fn],
env=self.env,
stderr=STDOUT)
except CalledProcessError as e:
msg = (
'Error running the command %s\n%s\nContents of file %s:\n\n%s'
% (
' '.join([interpreter, fn]),
'env=%s' % self.env,
fn,
'----\n%s\n----' % open(fn).read(),
))
ErrorClass = (FuturizeError
if 'futurize' in script else PasteurizeError)
raise ErrorClass(msg, e.returncode, e.cmd)
return output
secFiles = cms.untracked.vstring()
eventsToProcess = []
# Query DAS for a ROOT file for every lumisection
for ls in range(options.minLumi, options.maxLumi + 1):
if fnmatch.fnmatch(str(ls), options.lumiPattern):
read, sec = filesFromDASQuery("file run=%d dataset=%s lumi=%s" %
(options.runNumber, options.dataset, ls))
readFiles.extend(read)
secFiles.extend(sec)
# Get last eventsPerLumi of events in this file
command = "edmFileUtil --catalog file:/cvmfs/cms-ib.cern.ch/SITECONF/local/PhEDEx/storage.xml?protocol=xrootd --events %s | tail -n +9 | head -n -5 | awk '{ print $3 }'" % read[
0]
print(command)
events = subprocess.check_output(command, shell=True)
events = events.split(b'\n')
events = filter(lambda x: x != b"", events)
events = map(int, events)
events = sorted(events)
events = events[-options.eventsPerLumi:]
eventsToProcess.append("%s:%s:%s-%s:%s:%s" %
(options.runNumber, ls, events[0],
options.runNumber, ls, events[-1]))
eventRange = cms.untracked.VEventRange(eventsToProcess)
print("Got %d files." % len(readFiles))
source = cms.Source("PoolSource",
fileNames=readFiles,
